Abstract: In one embodiment, a method includes receiving training data, the training data including training video data representing video of a location in a quiescent state, training a neural network using the training data to obtain a plurality of metrics, receiving current data, the current data including current video data representing video of the location at a current time period, generating a reconstruction error based on the plurality of metrics and the current video data in the embedded space, and generating, in response to determining that the reconstruction error is greater than a threshold, a notification indicative of the location being in a non-quiescent state.

Abstract: In one embodiment, a device receives a network policy based upon, at least in part, a physical network, and configures a design of an industrial network overlay on the physical network based upon, at least in part, the network policy. The configuring, according to the techniques herein, may generally include: determining a number of virtual local area networks (VLANs) within the industrial network overlay; determining which devices of the physical network are on which VLAN; determining placement of at least one virtual firewall within the industrial network overlay; and determining at least one communication path for the industrial network overlay between at least two devices.

Abstract: A method comprises: obtaining, at an apparatus, first voice data from a first user device associated with a first speaker participant in a communication session; detecting voice data loss or degradation in the first voice data; determining whether prediction probability of correctly compensating for the voice data loss or degradation is greater than a predetermined probability threshold; if the prediction probability is greater than the predetermined probability threshold, first compensating for the voice data loss or degradation using historical voice data received by the apparatus prior to receiving of the first voice data, the first compensating producing first compensated voice data; if the prediction probability is not greater than the predetermined probability threshold, second compensating for the voice data loss or degradation by inserting noise to the first voice data to produce second compensated voice data; and outputting the first compensated voice data or the second compensated voice data.

Abstract: In one embodiment, a device in a network receives a set of sensor data from a plurality of sensors deployed in a location. The device determines a physical layout for furnishings in the location based on the received set of sensor data. One or more of the furnishings is equipped with one or more actuators configured to move the equipped furnishing in one or more directions. The device generates an instruction for the one or more actuators of a particular one of the furnishings based on the determined physical layout for the furnishings. The device sends the instruction to the one or more actuators of the particular furnishing, to implement the determined physical layout.

Abstract: In one embodiment, a device in a network receives information regarding one or more traffic flows in the network. The device identifies a particular one of the one or more traffic flows as a seasonal traffic flow based on the information regarding the one or more traffic flows. The device determines whether a service level agreement associated with the seasonal traffic flow is met. The device causes a time-based path for the seasonal traffic flow to be provisioned, in response to a determination that the service level agreement associated with the seasonal traffic flow is not met.

Abstract: In one implementation, sub-interfaces are defined in Layer three (L3) tunnels, such as generic routing encapsulization (GRE) or Internet protocol security (IPsec) tunnels. Sub-interfaces inside a L3 tunnel may be preferred to using several L3 tunnels. The flow label of the tunnel header is used to define sub-interfaces of a tunnel interface. The flow label is populated with a routing instance identifier to index the sub-interfaces.

Abstract: In one embodiment, a method comprises determining a movable network device is moving along a repeatable sequence of access point devices in a deterministic network; and establishing a deterministic track along the repeatable sequence of access point devices, the deterministic track comprising insertion slots allocated for insertion of a data packet, by the movable network device, into the deterministic track via any one of the respective access point devices.

Abstract: In one embodiment, a method comprises a network device identifying, in a time slotted network allocated timeslots for exclusive control of data transmissions with at least a second network device, a first schedule of first timeslots allocated for transmission and reception of packets having a first priority and a second schedule of second timeslots allocated for transmission and reception of packets having a second priority lower than the first priority, the second schedule overlapping the first schedule; and the network device shifting the second schedule of timeslots, relative to the first schedule, by a slot-frame shift (SFS) interval that causes a corresponding listen-before-talk interval in each of the second timeslots to be initiated coincident with or after transmission is enabled for any packet having the first priority.

Abstract: In one embodiment, a device in a network inserts a profile tag into an address request sent by an endpoint node in the network to a lookup service. The lookup service is configured to identify one or more addresses with which the endpoint node is authorized to communicate based on a profile for the endpoint node associated with the inserted profile tag. The device receives an address response sent from the lookup service to the endpoint node that indicates the set of one or more addresses with which the endpoint node is authorized to communicate. The device determines whether a communication between the endpoint node and a particular network address is authorized using the set of one or more addresses with which the endpoint node is authorized to communicate. The device blocks the communication based on a determination that the particular network address is not in the set of one or more addresses with which the endpoint node is authorized to communicate.

Abstract: In one embodiment, a method comprises identifying, by a path computation element, essential parent devices from a nonstoring destination oriented directed acyclic graph (DODAG) topology as dominating set members belonging to a dominating set; receiving, by the path computation element, an advertisement message specifying a first dominating set member having reachability to a second dominating set member, the reachability distinct from the nonstoring DODAG topology; and generating, by the path computation element based on the advertisement message, an optimized path for reaching a destination network device in the nonstoring DODAG topology via a selected sequence of dominating set members, the optimized path providing cut-through optimization across the nonstoring DODAG topology.

Abstract: Various implementations disclosed herein enable transforming mutable wireless coverage areas using network coverage vehicles (NVCs) that are orchestrated by a network coverage controller. In various implementations, the method includes receiving coverage area performance characterization values from NCVs configured to provide a plurality of mutable wireless coverage areas. In various implementations, an arrangement of the mutable wireless coverage areas mutably defines the service area, which changes in accordance with changes to the arrangement of the mutable wireless coverage areas. In various implementations, the method also includes determining NCV operation adjustments for some of the NCVs based on the received coverage area performance characterization values in accordance with a service performance metric; and, altering an arrangement of one or more of the plurality of mutable wireless coverage areas within the service area by providing the NCV operation adjustments to some of the NCVs.

Abstract: In one embodiment, a method comprises creating, in a computing network, a loop-free routing topology comprising a plurality of routing arcs for reaching a destination network node; identifying, within the loop-free routing topology, non-congruent paths for a source network node to reach the destination node; and determining, from the non-congruent paths, a non-congruent path pair providing no more than a prescribed difference of latency from the source network node to the destination node, enabling the source network node to forward a data packet in a first direction of the one non-congruent path pair and a bicasted copy of the data packet in a second direction of the one non-congruent path pair, for reception of the data packet and the bicasted copy by the destination node within the prescribed difference of latency.

Abstract: In one embodiment, a network node monitors communications between a sender node and an intermediary receiver node during a set of time slots of a channel hopping schedule. The sender node, intermediary receiver node, and a final destination node for the communications may all be located along a primary communication path in the network. The network node stores a copy of one of the communications sent from the sender node to the intermediary receiver node during a particular time slot in the set of time slots. The network node forwards the copy of the communication to a listener node configured to monitor communications between the intermediary receiver node and another node located along the primary communication path. The intermediary receiver node is also configured to monitor communications between the network node and the listener node.

Abstract: In one embodiment, a device in a wireless network receives a request from a node in the network requesting electrical power. The device determines one or more power transmission parameters for the node. The device determines a power transmission schedule for the node. The device sends wireless network communications to the node in response to the request and based on the determined one or more power transmission parameters and transmission schedule for the node. The node converts the wireless network communications into stored electrical power.

Abstract: In one embodiment, a device in a network receives packet arrival information for a packet from a neighbor of the device in the network. The packet arrival information indicates a likelihood of the packet being received by a target node that is moving in the network were the packet forwarded by the neighbor to the target node. The device forwards the packet to the target node based on a determination that the device has a higher likelihood of the packet being received by target node were the packet forwarded by the device to the target node than were the packet forwarded by the neighbor to the target node.

Abstract: In one embodiment, a method comprises promiscuously detecting, by a network device in a wireless data network having a tree-based topology for reaching a root device, a wireless data packet transmitted by a source network device and specifying a destination device in the wireless data network; determining, by the network device, that the destination device is within a first sub-topology provided by the network device to reach the root device, wherein the source network device is within a second distinct sub-topology provided by a parent device of the source network device to reach the root device; and causing installation of a bypass path, bypassing the root device, based on the network device generating and transmitting an instruction to the parent device to install a route entry causing a data packet destined for the destination device to be routed by the parent device directly to the network device.

Abstract: In one embodiment, a method comprises a first wireless network device identifying a deterministic receive slot reserved for reception of a first data packet from a second wireless network device along a deterministic track in a deterministic network; the first wireless network device transmitting an acknowledgement in the deterministic receive slot, to the second wireless network device, in response to successful reception of the first data packet in the deterministic receive slot; the first network device transmitting a second data packet in the deterministic receive slot, following the acknowledgement, to the second network device.

Abstract: In one embodiment, a method comprises: a root network device of a tree-based network topology identifying an instability in an identified child device attaching within the tree-based network topology; the root network device generating and storing in a routing information base table, for each sub-child device reachable via the identified child device, a corresponding source-route path starting with the identified child device and ending at the corresponding sub-child device; the root network device adding, to the routing information base table, a current path for reaching the identified child device, enabling the root network device to generate a path for reaching any one sub-child device using the corresponding source-route path via the current path of the identified child device.

Abstract: In one embodiment, a method includes receiving training data, the training data including training video data representing video of a location in a quiescent state, training a neural network using the training data to obtain a plurality of metrics, receiving current data, the current data including current video data representing video of the location at a current time period, generating a reconstruction error based on the plurality of metrics and the current video data in the embedded space, and generating, in response to determining that the reconstruction error is greater than a threshold, a notification indicative of the location being in a non-quiescent state.

Abstract: In one embodiment, a method includes training a deep neural network using a first set of network characteristics corresponding to a first time and a second set of network characteristics corresponding to a second time, generating, using the deep neural network, a predictive set of network characteristics corresponding to a future time, and assigning a task of a distributed application to a processing unit based on the predictive set of network characteristics.